Meh. Overreaction thy name is Dallas. First I have been given permission to put up a NY/A Graph weekly on my posts. So let's get right to it.
First O NY/A
Given the way averages work I shall assume, for the sake of argument, that league average is an 8-8 team. Because this is Offensive NY/A you want the red line to be above the green line for the season. (Two games is too small a sample size to be concerned yet.) YAY Our offense was league average yesterday. WOOOOOOOO.
To be fair the passing offense in the first half was a nice 6.6 NY/A good enough for 12th in the league!!! Well ok that number sucks, and football isn't played in halves. Not liking the Bill Callahan offense so far.
Game 1 was just terrible. Ouch. To be fair Game 2 was much better versus a team whose NY/A attempt on offense is a staggering... 4.9. Ouch We let them have better offensive production than their average on the season. Still small sample sizes right?
Again we want to be better than average on defense so we want the red line to be lower than the green line. Surprisingly I'm unenthused by the defense.
Next: NY/A Differential VS. League Average
I'm going to keep something I said in my last post as a useful record for how we ought to look at NY/A Differential:
20 of the 21 (95%) who scored a 1 or higher on their differential had a winning record
14 of 23 (61%) who scored between .1 and .9 had a winning record
0 of 3 (0%) who scored a 0 had a winning record
8 of 30 (27%) who scored between -.1 and .9 had a winning record
0 of 19 (0%) who scored -1 or lower had a winning record.
League Average is hovering right around where you'd expect considering that league average teams go 8-8. We on the other hand currently hover around -1.13 putting us - with an EXTREMELY small sample size - at a 0% shot to make the playoffs. On the other hand the game against KC was fairly nice. We had a positive 1.21 game versus them, which really means we played pretty well as a whole.
Which brings me to the last graph that I decided to add on a whim.
My aim with this set of graphs is to show that the Dallas Cowboys will most likely remain around league average in turnover margin over the course of the season. Stay tuned!
On to the next line of topic:
Turnover Margin. So I'm looking at a study that measures 23 weeks of NFL Football and I discovered something interesting. Did you know that teams that get 4 net or more turnovers a game win 100% of the time (or something absurdly high like that)? Did you know that teams that lose the turnover battle by 2 win 21.95% of the time? Interesting huh? Maybe this loss (much like the win against the Giants) has more to do with luck than actual skill? Just a thought.
Finally, I've read a pernicious line of thinking recently that has begun to aggravate me a little. "Any time the defense gives up 20 points or less in the game you should win that game." I don't even know how to handle this assertion because it is so baffling to me that I generally try to ignore it.
Consider two separate things
(1) Scoring is determined by situational play. Where you start with the football and where your opponent starts with the football greatly affects your ability to score it. If you're consistently having to drive 80 yards down the field for a score your PPG will be less than someone who only has to drive 20 yards down the field. Further the necessity to score is determined by the kind of game you're playing. Sometimes when you're ahead you take the foot off the gas to limit making a mistake. Therefore it doesn't seem like the volume stat points given up is going to say anything about your likelihood of winning the game.
(2) It just makes no sense from a statistical standpoint. There is literally no correlation between Points given up over the course of a season and winning a football game (.028) Think about that. Good teams are not the ones who give up the fewest points. They're the ones who make sure the differential between their point totals and their opponents point totals remains positive over the course of the season.
The defense didn't necessarily play well just because it held its opponents to less than 20 points. It played alright given it held its opponents to its season average. (I mean if you want to rejoice in that that's your prerogative) That number will either look a lot better or a lot worse as the season progresses.
Man small sample sizes take the fun out of everything.